starlink-fleet-data / README.md
juliensimon's picture
Update Starlink fleet: 10,169 satellites (8,827 operational, 1,297 raising, 44 deorbiting)
fe6d31b verified
metadata
license: cc-by-4.0
pretty_name: Starlink Constellation Fleet Data
language:
  - en
description: >-
  Daily health snapshots of the SpaceX Starlink constellation, derived from
  CelesTrak GP (General Perturbations) data. Tracks satellite count, orbital
  shells, operational status, and ISL capability acro
task_categories:
  - time-series-forecasting
  - tabular-classification
tags:
  - space
  - starlink
  - satellites
  - orbital-mechanics
  - tle
  - spacex
  - constellation
  - open-data
  - norad
  - leo
  - mega-constellation
  - tabular-data
  - parquet
size_categories:
  - 1K<n<10K
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/daily_snapshots.parquet
    default: true

Starlink Constellation Fleet Data

Orbital sunrise illuminating Earth's atmosphere, seen from the ISS

Credit: NASA

Part of a dataset collection on Hugging Face.

Dataset description

Daily health snapshots of the SpaceX Starlink constellation, derived from CelesTrak GP (General Perturbations) data. Tracks satellite count, orbital shells, operational status, and ISL capability across five inclination-based shells.

Starlink is the largest satellite constellation ever built, representing a fundamental shift in how broadband internet is delivered globally. SpaceX deploys satellites into five distinct orbital shells, each defined by its inclination and target altitude. The constellation operates in low Earth orbit (LEO) at altitudes between 328 km and 570 km, where orbital periods of roughly 90 minutes mean each satellite circles the Earth about 16 times per day.

Understanding constellation health requires tracking each satellite through its lifecycle: initial deployment to a parking orbit, orbit raising via Hall-effect ion thrusters, operational service at target altitude, and eventual controlled deorbit. The mean motion derivative serves as a reliable proxy for whether a satellite is actively thrusting upward or decaying. The inter-satellite laser link (ISL) capability, rolled out starting in 2022, enables direct satellite-to-satellite routing without ground relay.

This dataset is suitable for time-series forecasting, tabular classification tasks.

Schema

Column Type Description Sample Null %
date datetime64[ns] UTC date of the daily snapshot; one set of rows per date per shell 2019-05-25 00:00:00 0.0%
shell_id int64 Integer shell identifier (0-4); maps to inclination bands: 0=33 deg, 1=43 deg, 2=53 deg, 3=70 deg, 4=97.6 deg 2 0.0%
total_count int64 Total number of Starlink objects tracked in this shell on this date, including all statuses 3 0.0%
operational_count int64 Satellites with perigee altitude within the shell's operational band (typically 460-570 km depending on shell) 0 0.0%
raising_count int64 Satellites currently maneuvering toward their target shell altitude via Hall-effect ion thrusters 1 0.0%
deorbiting_count int64 Satellites in active controlled deorbit below their shell band with strong positive mean_motion_dot, or below 300 km 0 0.0%
reentered_count float64 0.0 6.1%
isl_operational_count int64 Operational satellites equipped with inter-satellite laser links (ISL); ISL-capable units were deployed from 2022 onward depending on shell 0 0.0%
avg_altitude float64 441.08666666666664 6.1%
min_altitude float64 439.58 6.1%
max_altitude float64 443.58 6.1%
new_launches int64 Reserved for future use; currently always 0 3 0.0%
anomalous_count float64 0.0 6.1%
shell_name str Human-readable shell label encoding inclination and target altitude, e.g. 'Shell 3 (53 deg / 550km)' Shell 3 (53° / 550km) 0.0%

Quick stats

  • 10,169 Starlink satellites tracked
  • 8,827 operational, 1,297 raising, 44 deorbiting
  • 9,205 ISL-capable satellites
  • 7,323 daily snapshot rows (2019-05-25 to 2026-04-21)

Usage

from datasets import load_dataset

ds = load_dataset("juliensimon/starlink-fleet-data", split="train")
df = ds.to_pandas()
from datasets import load_dataset

ds = load_dataset("juliensimon/starlink-fleet-data", "daily_snapshots", split="train")
df = ds.to_pandas()

# Constellation growth over time
growth = df.groupby("date")["operational_count"].sum()
print(growth.tail(10))

# Per-shell fill rates
latest = df[df["date"] == df["date"].max()]
for _, row in latest.iterrows():
    print(f"{row['shell_name']}: {row['operational_count']} / {row['total_count']}")

# Plot operational growth by shell
import matplotlib.pyplot as plt
for sid in sorted(df["shell_id"].unique()):
    shell = df[df["shell_id"] == sid]
    plt.plot(shell["date"], shell["operational_count"], label=shell["shell_name"].iloc[0])
plt.xlabel("Date")
plt.ylabel("Operational Satellites")
plt.title("Starlink Constellation Growth by Shell")
plt.legend()
plt.show()

Data source

https://celestrak.org/

Update schedule

Daily at 08:00 UTC via GitHub Actions

Related datasets

If you find this dataset useful, please consider giving it a like on Hugging Face. It helps others discover it.

About the author

Created by Julien Simon — AI Operating Partner at Fortino Capital. Part of the Space Datasets collection.

Citation

@dataset{starlink_fleet_data,
  title = {Starlink Constellation Fleet Data},
  author = {juliensimon},
  year = {2026},
  url = {https://huggingface.co/datasets/juliensimon/starlink-fleet-data},
  publisher = {Hugging Face}
}

License

CC-BY-4.0